ABSTRACT: Although oral anticoagulant (OAC) therapy provides superior stroke risk reduction in patients with Atrial Fibrillation (AF), it is widely underutilized with hemorrhage being a major deterrent. OAC related hemorrhage accounts for 33.3% of adverse-drug- related hospitalizations in the US and is a critical barrier to institution of therapy. This revised application builds on our successful project identifying the influence of genes on warfarin dose, anticoagulation control, and hemorrhage (n=1310; 43% Black). We expand our efforts to incorporate new OACs to identify predictors of hemorrhage in Dabigatran (DBG; n=500) and warfarin (n=1000; 590 accrued) treated AF patients through four specific aims. Aim 1 will elucidate the influence of common and rare genetic variation on risk of warfarin-related hemorrhage using a genome-wide approach. The discovery efforts will be grounded in 700 warfarin-related hemorrhage case-control pairs with replication in an independent prospective cohort of 1000 warfarin-treated AF patients. Aim 2 will elucidate the influence of race, kidney impairment and concurrent antiplatelet therapy on risk of warfarin-related hemorrhage in the prospective cohort of 1000 warfarin-treated AF patients. Aim 3 will elucidate the influence of kidney impairment and concurrent antiplatelet therapy on risk of warfarin- related hemorrhage in the prospective cohort of 500 DBG-treated AF patients. Aim 4 will incorporate patient-specific genetic and clinical factors into refining (for warfarin) and building (for DBG) clinical prediction rules (CPRs) to personalize the prediction of hemorrhage. The AF patient-cohort will provide a robust foundation for future efforts that will incorporate other new OACs namely rivaroxaban and apixaban. The focus on AF lays the foundation for future real-world comparative- effectiveness evaluation in a population representative of clinical practice.